Policy proposals guided by empirical analysis of the negative effects of policing overlook the indirect and virtual harms of police encounters. These harms produce negative health outcomes in highly policed communities that extend far beyond direct police encounters. Affective analysis is the only way to grasp this dimension of police brutality.
I. The Neighbourhood Effect of Policing According to recent studies, the impact of aggressive policing tactics extends beyond the individual who experiences a direct encounter with police officers. Unsurprisingly, the individual’s social and familial circles are also impacted by such an encounter. Perhaps less intuitively, these studies also reveal how populations living in proximity to the areas where these tactics are deployed experience measurable decreases in health, many of which can be associated with stress and trauma.Such studies provide valuable data for establishing the ways police practices have effects on communities rather than simply individuals, and must therefore be analysed from a broader perspective.If populations who are merely in the same neighbourhood as aggressive police practices suffer, then it is insufficient to simply use police department statistics – usually stops, citations and arrests – to quantify negative health outcomes. The neighbourhood effect of policing policy is an important supplement to research at the individual scale. For example, Abigail Sewell and Kevin Jefferson analyse data collected under the New York City Stop, Question and Frisk (SQF) policy between 2009 and 2012 (Sewell and Jefferson 2016). The data they collect and analyse suggests that invasive pedestrian stops have a negative impact not only on those individuals who are stopped, but also on populations who have not themselves been stopped, but merely live in proximity to police stops. In neighbourhoods where stops are more likely, mental and physical health outcomes are worse; populations in these neighbourhoods suffer disproportionately high instances of diabetes, hypertension, insomnia and obesity. In places where there is a higher likelihood that force will be used in the course of an encounter, negative health disparities are more severe. The authors draw a correlation between this more aggressive and invasive police tactic and the health outcomes suffered by populations who live in the targeted area. While they invite additional study to confirm their findings, they conclude that their evidence suggests that the consequences of the institutionalization of the carceral state are far-reaching’ (Sewell and Jefferson 2016). Sewell et al.’s research extends previous studies that suggest a strong association between psychological distress and neighbourhood-level use- of-force and SQF rates, particularly as it pertains to males of colour. Because males are more often subject to these encounters, the authors suggest that the stress (and related set of health outcomes) is likely to be a result of the higher likelihood of SQF.The fact that these health outcomes are found most acutely in male populations, where there is a higher likelihood of being stopped by police officers and a higher likelihood of a violent encounter with police officers, suggests that proximity alone is not the explanatory factor (Sewell et al. 2016). Thus, Sewell and colleagues show that certain populations experience the effects of policing policy even when they do not directly encounter police officers. Their research better outlines the expanded impacts of policing policy in two important ways. First, there appear to be discoverable effects from encounters that never take place.The neighbourhood effect reveals the impacts of a non-event: the fact that an encounter does not occur does not occlude a real impact of the non-encounter–an impact that is entangled with an actual encounter that does not take place. The non-event and the event2 are entangled. The very risk of an aggressive encounter is enough to yield measurable, observable results. Such research accounts for the impacts of something that does not occur – but could and in other cases does occur – within a given neighbourhood and as a result of a particular practice. Second, a specific subset of police practices – SQF, surveillance, brutality, invasive searches, aggressive stopping practices, profiling and others – appears to have an ever-widening ripple effect on populations who do not directly experience such practices. Some of these ripples can be quantified and analysed when we move from the individual to the neighbourhood and beyond. Thus, Sewell et al. make a strong case for the benefits of including a larger set of criteria in the study of public and policing policy. However, there are methodological limitations to the social scientific methodsused in this type of study that make it difficult (or perhaps impossible) to fully capture the scale and intensity of such police practices.3 The confines of this scientific approach must necessarily exclude elements that are not susceptible to calculable, repeatable, data- driven analyses. The method must eliminate variables that interfere with significant results derived from its data, and it must designate the boundaries of the data in order to calculate findings. Each hypothesis must be testable (and perhaps replicable) if it is to be meaningful. Human behaviour must be translated into discrete data points (for example, convert self-reported emotions into a numerical scale), calculated using external proxies (for example, aggregate health outcomes or income levels in different spaces or over different periods of time), or speculatively deduced through observation (for example, measure micro- reactions to simulated experiences in real time). Such analysis can and does yield important, meaningful results to a wide variety of research questions, but it also excludes a realm of empirics that is incompatible with these methods. Because no single view of empirical study is accepted as comprehensively and universally objective, the next section analyses the theoretical and methodological resources provided in Difference and Repetition to identify limitations in social scientific methods and propose a complementary set of considerations that may yield a richer account of policy effects. More broadly, I argue that basing research and policy on such a narrow conception of empirics often fails to account for elements of the ‘real’ that have a significant impact on its explanatory effectiveness. Such limitations are most acute in cases where scientific research fails to recognise the fundamental limitations of its methodological parameters. In the case of studies on the neighbourhood effect, there is a risk of interpreting human behaviour and community relations as something less than the complicated, complex, fluctuating fields of ideasthat Deleuze indicates them to be. How much is lost when these complicated fields are reduced to positivist analysis? The confluence of this approach to research (and the urban policing policies often derived from it) runs counter to the normative aspirations with which I conclude the essay. In contrast, a model of analysis that provides space for the other to remain in a state of becoming–even on terrain that is outside the confines of social scientific research methods–is more likely to foster and account for the creative process of becoming while calling for a less invasive set of policing policies.
II. Transcending the Neighbourhood Effect A. The Problem of Scaling and Localisability
The world that social scientists observe and study is the realm of what Deleuze calls the actual: the recognisable world experienced on a daily basis. However, the ‘actual’ is not nearly sufficient for characterising the world in its emergent and creative complexity. For Deleuze, the actual emerges from a far richer realm that cannot be directly accessed or recognised: the realm of the virtual. The world as we experience it emerges from the virtual, but it bears no relation to the realm of the actual. Prior to the actualised space of objects, concepts and bodies is a realm of Ideas that is yet to be actualised. Following Kant, Deleuze argues that these Ideas can be understood as ‘essentially problematic’ while ‘true problems are Ideas’ (Deleuze 1994: 168). Problems themselves ‘are the real objects of Idea’, even though those objects remain ‘outside experience’ and ‘represented only in problematic forms’ (Deleuze 1994: 169). True problems are not safes to crack or codes to break; they bear no relation to solutions. Instead, problems–and therefore Ideas–are multiplicities of non-reducible differential elements that cannot be divided, subtracted from or resolved. (Deleuze 1994: 182) The various elements in virtuality that constitute the Idea lack both a ‘sensible form’ (which would be found only in an actualised form) and a ‘conceptual signification’ (which would be soluble or recognisable, and therefore not a problem); Ideas are ‘on the way’ to being actualised as a result of an unknown and dramatic triggering event or change of conditions (Deleuze 1994: 182–3). Actualisation is the creative process of moving from the virtual to the actual. Deleuze argues that all objects are the result of a movement from a realm of Ideas that exists in a pre-recognisable space of completely differentiated elements; they exist only as pure potentialities. These elements are intermittently drawn together into a series of points that constitute a structure, after which something begins to come together. Although it may be sensed as an emergent formation, it cannot yet be recognised as an object with extensity. At some trigger point, in a dramatic moment of emergence, something coagulates into extensity. The emerging formation does not resemble its previous stages, though it is a result of them (Deleuze 1994: 212). Instead, what crystallises into the object we now encounter could not have been anticipated or described in advance. Both the timing of the appearance and the qualities of the object are surprising. The differential elements that came together in the realm of the virtual are now ‘cancelled out’ in the realm of actualisation; Deleuze argues that the object is now ‘differenciated’.4 It can be recognised, described, compared, identified, quantified and categorised–it has now entered the realm of the actual. It may be tempting to consider only the realm of the actual as ‘real’ because it is constituted by objects that are readily accessible, but Deleuze insists that the virtual is equally ‘real’ (Deleuze 1994: 208). The virtual is the genetic principle from which the actual derives; the actual cannot exist without it. Although the realm of the virtual is not contained by space or linear time, it does have an impact on both space and time, and is thus real. The virtual is in oppositional relation to the actual – not the real – because the actual realm and the virtual realm bear no relation to one another. Both the actual and the virtual are real, however, because we experience the world only as a result of the virtual actualising.5 In many social scientific studies, including the ones mentioned here, social scientists identify and study a territorially bound space (or set of spaces) in order to gather a discrete data set to analyse. Sewell et al., for example, gather data in a defined set of New York City locations over a specific period of time. These boundaries are necessary to provide a finite scope of analysis. If significant, their findings can then be extrapolated based on this specific location or applied to other contexts based on the findings of the initial study. However, any social scientific approach that is confined to a definable space over a specified period of time must necessarily exclude the considerations of the virtual for reaching its conclusions. It must rely instead on the realm of the already- actualised for its data and analysis. Such a limitation is significant because the virtual–that rich terrain of potentiality from which all of extensity is actualised–can be neither cordoned off geographically and chronologically, nor ignored. Rather, it must be included and thought through. To illustrate how the virtual influences one’s life, even while eluding localisability, consider an individual viewing a video of a police shooting. The images of police officers unjustly shooting an unarmed individual may elicit an affectively charged response regardless of whether the shooting took place in one’s neighbourhood or across the world. Both the shooting and the observation of the shooting are examples of complex processes of actualisation. The person viewing the video may experience an impact of the image as they speculate an analogous possibility occurring to them, in their neighbourhood, in an encounter with their local police force. Or, a viewer may simply bear witness to the violence being carried out and experience a real impact from its observance, whether such an encounter is likely to occur to the observer or not. One’s personal connection to the individuals in the video is not necessary for one to incur a variety of real effects from having seen it, and the nature of those impacts cannot be anticipated in advance.Policing policies may not be consistent or generalisable from one neighbourhood to the next, but the effects of policing activity in one area may reach into geographically distant spaces and have an effect nonetheless. The intensities beneath the register of recognisability travel and expand, regenerating their own force of impact. These intensities inhabit the relation between the viewer, the video, their lived experiences and the milieu in which they find themselves. Reactions to observed violence may not be consistent from one person to the next, or from one population to the next; they may vary without apparent explanation.Because the world is derived from the rich terrain of the virtual, experiences and milieus may coalesce to impel an unexpected reaction to a particular video, news story, rumour, conversation, unrelated source of frustration, policy, event, encounter, near-encounter or non-encounter. It is not possible to anticipate in advance where the effects of an event will occur; one cannot pin down the exact source of a particular reaction; one cannot draw a direct correlation or causation between an event or policy and those who are impacted by it;one cannot say in advance how acute or widespread a reaction to a stimulus will be. Because the process of actualisation is creative and cannot be fully accessed, the scale of an actualised effect cannot be calculated or predicted based on severity or frequency of occurrence. In most cases, a determination of cause is speculatively identified after the fact, based on retroactive analysis. Even in these cases, the complicated sprawl of contextual causes tends to be siphoned away in order to make an analysis ‘meaningful’ or ‘significant’. For example, it is impossible to definitively determine why the deaths of Freddie Gray and Michael Brown caused widespread political uprising, while the deaths of Terence Crutcher or Stephon Clark did not.Deleuze writes that the expression of the other presents to us a potential reality in the process of creatively coming to be. ‘Consider a terrified face’, Deleuze writes, when the observer of a terrified face does not know the source of terror. ‘This face expresses a possible world: the terrifying world.’The world expressed may not resemble the actualised world of extensity. Instead, it reflects ‘that state of the implicated or enveloped in its very heterogeneity with what envelops it: the terrified face does not resemble what terrifies it, it envelops a state of the terrifying world’ (Deleuze 1994: 260). When one encounters the terrified face of the other, they are suddenly implicated in a heterogeneous system; the radical separation between the subject and the other dissolves.The separation between the event and the reaction to the event are no longer at odds with one another. They envelop one another – the reality of the present becomes subsumed by the emergent possibility expressed by the other as it coalesces to constitute a new present moment. The present moment is a complicated multiplicity of possible worlds expressed by the other that now subsume my own world. A ‘virtual’ experience of the police shooting is real when it is experienced through its actual viewing, but the virtual and the actual do not remain in discrete realms. Instead, the virtual reaches into the actual and expresses a world in which both the event and the non-event are ‘real’. The world in which there is no shooting is subsumed by the world in which there is a shooting, and this subsumption constitutes the present moment for the viewer. Deleuze’s understanding of the virtual’s relation to the actual is a resource for thinking through concepts and events involving non- localisable, non-scalable processes. Again, the process of moving from the realm of virtual to material realities that we can recognise is marked by creativity. The exact moment and nature of an event cannot be reliably predicted in advance; a moment of crystallisation may appear disproportionate to the conditions that led to it, and the result may not bear any resemblance to the trajectory that preceded it. The realm of ideas that serve as the basis for actualisation cannot be directly grasped, observed or measured (Deleuze 1994: 183). If the cause is intensive rather than extensive, and the effect is the result of the virtual actualising, then both the source and result of the event cannot be geographically pinpointed. The conditions from which an event derives may not occupy a defined space, or they may not be limited to that defined space. The scientific practice of defining a space, time and scale from which to derive and analyse data fails to account for the creative process of actualisation as described by Deleuze. Even if the scale or effect of an event makes itself known in the realm of extensity, as is the case in Sewell’s study, that scale is not reducible to those observable, actualised effects. Including Deleuze’s resources in an analysis of policing policy adds richness to the way policing policies and their effects are understood, even though (and because) they cannot be localised, and even though the scale of effects cannot be predicted or comprehensively measured.
B. The Problem of Affect and Its Resistance to Social Scientific Interrogation
Henri Bergson’s conception of affect, from which Deleuze derives much of his own theory, is a helpful supplement to the social scientific approach to studying aggressive police tactics. However, it also poses a serious critique of it. Following Bergson, I take affect to be that which precedes emotion–a disorganised, pre-emotional proto-feeling.6 A discernible and identifiable emotion is the actualised version of the affective sense that preceded it; affect is the virtual or genetic component of emotion. An analysis of the affective side of policinghelps elucidate the empirical findings of Sewell et al. and others, but it also reveals how certain aspects of human behaviour are less susceptible to social scientific study. For populations who live in neighbourhoods targeted by more aggressive policing tactics–in particular, black males who are more likely to be experience a police encounter (Sewell et al. 2016)–there is an affective charge associated with the possibility of police contact.Speculatively, this charge may derive from multiple sources, including the potential danger of brutality or lethal force used on black males during police encounters; a sense of powerlessness in relation to police departments who may appear to act with impunity; a sense of inequality based around the disparity between targeted and non-targeted neighbourhoods or populations subject to such tactics; a perception of lawlessness surrounding the rules governing police behaviour and the legal consequences resulting from misconduct; the apparent racial disparity in legal accountability after an incident of police misconduct; the perception of arbitrary or racially motivated selection for pedestrian stops; the continual and often futile effort to avoid being perceived or portrayed as a dangerous individual in the eyes of the police or society more generally. For Deleuzians, such a charge is further complicated because the source and effect of the affective sensation are tangled up with one another: affective forces inform these beliefs as they populate the cognitive portions of our thought processes, but also are the result of these informed beliefs(Connolly 2007). Affect cuts across the intellectual threshold; it precedes it, co-constitutes it and emerges from it. It is not sufficient to say that one has an idea of what might occur in one of these encounters, or even to say that one is scared of what might result from a police encounter–those thoughts and emotions are imbricated with the affective sensations that precede them (and then reorganise them). It would be more accurate to say one’s behaviour is pre-organised and directed by an affective experience of the world. That experience is antecedent to the cognitive sense-making of one’s experience that one undertakes. Additionally, the relationship between the two cannot be clearly traced–they can be separated only in conceptual terms. Affect and cognition are examples of the virtual and the actual; the cognitive component emerges from an affective sense, even if it bears little resemblance to its virtual origin. Because I take affect to be virtual rather than actual, social scientific methods struggle to analyse it; approaches that are confined to a world of spatial extensity fail to account for the way affect influences and constitutes the realm of the actual.Affect is real even though it is not possible to locate or map its precise functioning. Again, the realm of extensity is constituted by the intensive realm of virtuality:affect influences our behaviour, informs the way we understand the world, and modifies how we react to external stimuli. And, it cannot be isolated in an event, space, time or even the potential of an event in a particular space at a particular time. It cannot be placed within the subject’s mind or body. However, affect cannot be conceptualised without both the subject and the event(s). Affect is thus what Jairus Grove (2016) calls ‘semi-autonomous’. It is constituted through a relation between a subject (or group of subjects) and the potentiality of an event (a police encounter, in this case).Quantifying the impacts of this semi- autonomous element offers a glimpse at the actualised effect of affect, but does not – and cannot – offer a picture of affect as such. The observable world–the one that can be scientifically broken into data sets – is merely the end of a creative process of actualisation. We cannot access the realm of intensity directly; it can be sensed but not measured. Even if it were possible to designate and quantify the magnitude of an affective sensation, the relation between affect and its actualised result cannot be predicted. There is not reliable correlation between particular affective responses and a consistent behaviour that follows. Even if the considerable role affect plays is granted by positivists, the precise ways that affect organises our behaviour cannot be methodologically identified. Affect is thus a crucial and constitutive element of human behaviour, but cannot be located, observed, measured, quantified or correlated with behaviour in any reliable way. The extensive world that results from the virtuality of affect may be susceptible to social scientific study, but it provides a severely limited glimpse of the complicated nature of its genesis. Sewell et al.’s scope of analysis is and must be limited to the realm of the extensive and observable, and thus cannot scientifically incorporate affect as a mathematical variable. For example, the visceral experience of fear (or a less organised sensation that later coalesces into a recognisable emotion of fear) may result in measurable consequences that can be sorted into a data set, but the intensity and extent of these consequences are not reducible to discrete quantitative analyses. Although we could develop qualitative data by asking populations to rate their fear of a police encounter on a numerical scale, or gather quantitative data on the health outcomes of populations who live in SQF neighbourhoods, this data does not capture the intensive power of affect or even the myriad of untraceable impacts that can be connected with an affect-based lived experience. To understand why an encounter with the police is real whether it takes place in actuality or not requires the inclusion of a virtual world that is largely incompatible with social scientific methodology.It is possible to theorise how the virtual – the non-material – reaches into the world of the actual through an intensive relation and has an impact on behaviour, but it cannot be reliably measured. Such an effect is real but not actual; quantifying the observable results fails to capture the role of the virtual. Thus, Sewell et al.’s studies reflect an important consequence of affective relations, but they do not (and probably cannot) capture the way affect precedes, eludes and confounds empirical study.Put differently, Deleuze directs our attention to a realm that is not yet present in any recognisable way. Philosophy, he argues, cannot be interested solely in the already constituted world, because to do so is to exclude a much richer and more dynamic realm of virtuality. Deleuze criticises the uprightness of thought that emerges from the concordance of senses, in which all things are recognised and representable7 (Deleuze 1994: 133). But, by the very nature of its methodology, social science struggles with a world that cannot be made recognisable and analysable from a positivist perspective. The world of extensity can be measured, but it does not ‘bear any resemblance’ to the world of the virtual. Aggressive policing, surveillance, mass incarceration, red-lining, stop, question and frisk are all differenciated extensities. They are the inverted image of difference.8 To understand how this constellation of concepts operates on their fundamental scales, we must turn towards the virtual. In the case of Sewell’s studies, one can observe and record recognisable effects on a population’s health as a result of particular policing practices. However, a rich terrain of potentially unrecognisable or non-calculable impacts cannot be analysed using these tools. Deleuze’s understanding of intensive affect provides a framework for thinking through the less recognisable or more obscure aspects of an actualised practice.